A director comes out of a board meeting with a new target: thirty percent of the team's work done with AI by the end of the year. Nobody in the room has said what that means, which work, or how it will be measured. She has to walk back to her team and make it make sense. She's a composite of leaders I've coached.
If you lead people in the Gulf right now, some version of that meeting has probably happened to you.
Why the pressure and the results don't match
Investment is enormous. Results, so far, are uneven. MIT's Project NANDA, in its 2025 report The GenAI Divide, found that despite 30 to 40 billion dollars of enterprise spending on generative AI, 95 percent of organisations in its research saw no measurable return. The small group that did get value had built the tools into how work is actually done, and used tools that learn from feedback and fit the way the work is done.
Meanwhile, individual people are often ahead of their organisations. PwC's Middle East survey found three quarters of employees in the region had used AI tools at work in the past year. Usually the problem isn't the people but how the work, decisions and trust around them are organised.
What helps, in practice
- Translate the target into work. Sit with your team and list the tasks that take the most time and add the least value. Start there, not with the tool.
- Make it safe to report what doesn't work. If people think admitting a failed experiment will cost them, you'll get glossy updates and no learning.
- Be honest about jobs. People will assume the worst if you say nothing. You may not have every answer, but you can say what you know, what you don't, and when you'll know more.
- Go back to the CEO with a better question. Instead of "how much AI", agree on "which outcome", such as faster hiring or fewer errors in a process. It's easier to deliver and easier to defend.
The leader's part
The tool can draft the report. It can't decide which risk is acceptable, hold the room when people are anxious, or take responsibility when an experiment goes wrong. Those are still yours. The leaders who do well with AI are usually not the most technical. They're the ones their teams trust enough to try new things in front of.
Questions to take back to your team
- Which hour of your week would you most like a machine to take?
- What are you worried about that you haven't said out loud?
- What would we need to see in three months to know this is working?
I coach senior leaders and HR leaders who are carrying this pressure. If that's you, book a free 30-minute call, or book a single session on one specific issue.